In order for a test construct to be valid, it needs to be compared to a predefined rubric for each specific test; but it is here that “the (construct) theory itself is not called into question: it is taken for granted. The issue is whether the test is a successful operationalisation of the theory”